Maggie's FarmWe are a commune of inquiring, skeptical, politically centrist, capitalist, anglophile, traditionalist New England Yankee humans, humanoids, and animals with many interests beyond and above politics. Each of us has had a high-school education (or GED), but all had ADD so didn't pay attention very well, especially the dogs. Each one of us does "try my best to be just like I am," and none of us enjoys working for others, including for Maggie, from whom we receive neither a nickel nor a dime. Freedom from nags, cranks, government, do-gooders, control-freaks and idiots is all that we ask for. |
Our Recent Essays Behind the Front Page
Categories
QuicksearchLinks
Blog Administration |
Wednesday, October 20. 2010Lies, Damned Lies, and Medical ScienceFrom an article about the wonderful Dr. John Ioannidis of the above title in The Atlantic (my bolds):
Clinical research always must be taken with a grain of salt, and today's "best practices" will be tomorrow's worst. The general press is utterly incompetent at evaluating such studies. I think they just grab at potential headlines, eg Study: Broccoli I know plenty of folks who have been told to "Take it - studies say it might help, and won't hurt." Who knows? I find it amusing to think that today we are no longer certain of a direct relationship between cholesterol levels and heart disease.
Posted by Dr. Joy Bliss
in Medical, Our Essays, Psychology, and Dr. Bliss
at
15:57
| Comments (10)
| Trackbacks (0)
Trackbacks
Trackback specific URI for this entry
No Trackbacks
Comments
Display comments as
(Linear | Threaded)
The root cause of this is the innumeracy of the collective media. They have no idea how scientific testing is conducted much less such things like multivariate statistical analysis.
For instance, the claim that Drug Y is 50% more effective than Drug X fails to note that the actual effectiveness of Drug X is 1% and 50% more effectiveness is statistically meaningless or that there might be much more effective treatments than Drug X or maybe they even tested Drug Y against common table salt or some other non-medicinal substance. What results do you want? I can deliver them! The peer review process was originally intended to catch this form of scientific chicanery but has become largely meaningless. All peers are dependent on government largess, so no one is willing to say a paper is fit only for the shredder when they know the author will soon be reviewing one of their papers. Plus, it is fairly easy to get two janitors and a secretary to sign off on a peer review form before submission. Does anyone remember when Fleischmann & Pons claimed to have developed "Cold Fusion"? Other experimenters pretty much tore them apart in months. This was the way science is supposed to work. It may not be comfortable but good research does get done. This whole who to trust thing is getting annoying.
All I know is that 9 out of ten doctors recommend Anacin. And 87% of scientists support the AGW theory- one activist told me that recently, in all sincerity.
And that Maxwell is my cup of tea. Coffee is a good example, because over the years we have learned that it is bad and that it is good for you. I came to the conclusion years ago that doctors do not understand basic statistics theory. They took a course in biostatistics and learned how to use all these statistical methods but they don't really understand the theory. For instance, you will always see a confidence interval in these medical studies. This interval is presented as setting upper and lower bounds on the parameter of interest, but it actually doesn't do that at all. Their calculated confidence interval is meaningless. The confidence interval has to do with repeated experiments. If you perform the same experiment an infinite number of times and calculate a confidence interval each time, then 95% of the confidence intervals (P=.05) will contain the parameter of interest. If you calculate the confidence interval once, all you can honestly say is that the parameter is either inside or outside the interval. You can't honestly claim there is a 95% probability the parameter is inside the interval.
Science is hard. Let's be real. Medical science has already picked the low hanging fruit. People live much longer because they don't die from smallpox, cholera, tuberculosis and a whole litany of other diseases that killed our ancestors. Our food, air and water are better. Modern sanitation has been a boon to civilization. I think Dr. Ioannidis fails to realize that what he's seeing is a feature, not a bug. I can't comment on the state of peer review, but it seems to this layman that the more experiments the better. Much of scientific progress is based on learning what's not true, on failure.
I've long held that the cholesterol causes heart disease theory was bunk. A cynic would note that it has generated a lot of unnecessary spending by an anxious public on medical testing, "remedial" drugs, and changes in diet.
One big problem with many of these medical shock stories is that corrolation is not causation. Add that they have very limited sample sizes and scopes and they are meaningless except for 15 minutes of MSM fame. A classic is the relationship of low level radiation exposure and cancer. We do have a legitamite data on HIGH doses, the ones that make you acutely sick. But for very low levels, like that for neighbors of nuclear power plants, there is NO data of statistical significance. However, our regulators use a very conservative assumption called the "linear effects hypothesis" which draws a straight line between known effects at high doses (like Hiroshima and Nagasaki) and zero dose. Of course NOBODY gets zero radiation dose. Our bodies are naturally radioactive, the soil contains radioactive isotopes and cosmic rays always come from the sky. Yet dopes like Harry Reid will argue that haul spent nuclear fuel in thickly shielded containers over the public highways or on railroads will cause a calculated number of cancers in innocent people. I once heard Reid's technical assistant once claim that nuclear power has killed 40,000 people in the US alone. I raised my hand and asked "Where are they buried? That's a a lot of carrion - are there mass graves and a memorial somewhere?" A classic is the relationship of low level radiation exposure and cancer. We do have a legitamite data on HIGH doses, the ones that make you acutely sick. But for very low levels, like that for neighbors of nuclear power plants, there is NO data of statistical significance. However, our regulators use a very conservative assumption called the "linear effects hypothesis" which draws a straight line between known effects at high doses (like Hiroshima and Nagasaki) and zero dose.
I have read in recent years a summary of some study which claimed that at a certain distance from the Hiroshima/Nagasaki blasts life expectancy was actually higher than normal: at a certain low level, radiation could add years to life. Ever been to Utah? Ra-di-a-tion. Yes, indeed. You hear the most outrageous lies about it. Half-baked goggle-box do-gooders telling everybody it's bad for you. Pernicious nonsense. Everybody could stand a hundred chest X-rays a year. They ought to have them, too
(Repo Man) "... help fend off Alzheimer’s disease" reminded me - a week ago I read an article (it had no cites, and I doubt I could find any since I do not pay for medical or Lexis/Nexis or whatever search engines, but...) that mentioned a couple of studies had shown something that delayed onset in the at-risk by 50 percent or more. Not that the author thought it would be wise to do that something - smoking cigarettes.
Change the science from medicine to climate.
It all fits into the same 5 pound sack of .... |